Journal
IEEE TRANSACTIONS ON MEDICAL IMAGING
Volume 40, Issue 12, Pages 3369-3378Publisher
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TMI.2021.3084748
Keywords
Image segmentation; Task analysis; Tumors; Training; Biomedical imaging; Neural networks; Feature extraction; Artificial intelligence; cancer detection; neural networks; regularization; residual learning; segmentation
Categories
Funding
- Korea Institute of Science and Technology [2E31122 2K02540, 2E31071]
- National Cancer Institute [1R01CA227713]
- Google Faculty Research Award the Stanford Bio-X Bowes Graduate Student Fellowship
- Human-Centered Artificial Intelligence of Stanford University
Ask authors/readers for more resources
The study introduces a new neural network architecture utilizing multiple output channels and implementing consistency regularization through residual learning to improve image segmentation performance. Validation on public data shows significant improvement in tumor detection and delineation, with potential broad applications in various deep learning problems.
Deep learning is becoming an indispensable tool for imaging applications, such as image segmentation, classification, and detection. In this work, we reformulate a standard deep learning problem into a new neural network architecture with multi-output channels, which reflects different facets of the objective, and apply the deep neural network to improve the performance of image segmentation. By adding one or more interrelated auxiliary-output channels, we impose an effective consistency regularization for the main task of pixelated classification (i.e., image segmentation). Specifically, multi-output-channel consistency regularization is realized by residual learning via additive paths that connect main-output channel and auxiliary-output channels in the network. The method is evaluated on the detection and delineation of lung and liver tumors with public data. The results clearly show that multi-output-channel consistency implemented by residual learning improves the standard deep neural network. The proposed framework is quite broad and should find widespread applications in various deep learning problems.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available